Location recognition system using ultrasonic waves

The location recognition system uses ultrasonic signals to identify the location of a target within an enclosed area such as a room. The approach was designed to reduce blind spots in monitored areas without the use of expensive systems employing several cameras with panning capability. The results obtained in the research may be useful in the development of simpler and cheaper human intrusion monitoring systems that could find various applications in security installations. Using relative amplitudes of ultrasonic signals received by eights receivers inside a test area, the system was able to show that it is possible to locate a target in an enclosed area with the help of a pattern recognition algorithm, even when occlusions are introduced. An algorithm was developed using Artificial Neural Network (ANN) techniques. The system went through two phases: the training phase and the operational phase. During the training phase, the data gathered from the receivers were stored in a database and were used as training data for the ANN. In the operational phase, the system used the trained ANN in order to locate a target in the area. The output is a display showing a pair of X and Y coordinates indicating the location of the target. The program displayed a four by four grid, representing 16 possible location coordinates on the test area. Further, tests showed that eight receivers and four transmitters are the minimum requirements for the system to be able to locate targets with acceptable accuracy. In this case, the accuracy percentage reached more than 80%, with or without an occlusion